Abstract: In response to the call of "carbon peak,carbon neutral",a large number of distributed energy（DG）are connected to the distribution network. This paper proposes a dynamic reactive power optimization strategy based on demand side response considering DG output and load dynamic change. In this strategy,the gray relational mapping method is used to divide the time periods,and different types of variables are optimized in the same time period. Then,the shunt capacitor bank（SCB）and on-line tap changer（OLTC）is fixed in multiple time periods,and the secondary static reactive power optimization correction is carried out. According to the characteristics of the model,different types of variables are solved by hybrid co-evolution algorithm,and an improved sparrow algorithm based on tent chaotic map and Levy flight strategy is proposed to improve the efficiency of solution. The simulation results show that the proposed strategy considering new energy output and demand response can optimize the load characteristics and obtain satisfactory reactive power optimization results while reducing the solution scale,moreover the hybrid co-evolution algorithm has certain advantages in solving mixed integer non convex nonlinear optimization problems.